---
datasets:
- irodkin/celeba_with_llava_shorter_captions
---
**For this model the only one available contol input is the conditioning_image from dataset irodkin/celeba_with_llava_shorter_captions** 

And this image the model actually ignores. Its just experimental text2image